work plan for soil remediation, sampling and analysis plan, site ...

6.0 DATA ASSESSMENT PROCEDURESReliability in analytical determination is maintained through strict adherence to quality controlprocedures. Procedures are designed to control both the accuracy and precision of analyticalresults. Depending on the level of certification of the data, a known method spike is routinelyanalyzed to ensure the accuracy of results. The procedure is to run the standard QA/QC andsample analysis with each lot of samples sent to the laboratory. If more than ten individualanalyses are made, additional standards will be analyzed at a rate of one standard per tenanalyses. Some procedures call for the use of either a surrogate spike or the standard additionof a known quantity of the analyte to a split of the sample being analyzed.Control charts will be prepared using an estimate of the spike recovery obtained from theliterature or determined by repeated analyses run in the laboratory. Each time the analyst runsa method spike, the results are entered on the control table. If a standard addition technique isused, a plot of instrument response versus added analyte concentration is made in order todetermine analyte concentration in the original sample. These are further explained in thelaboratory’s QAM.Replicate analyses will be performed on at least 10 percent of the samples processed by thelaboratory. A record of the precision of most analyses is kept by calculating and plotting theindustrial statistic I (which is equivalent to the coefficient of variation). Blanks are also runwith each batch of samples or individual sample analyzed regardless of the level of certificationof the data.The purpose of spikes, blanks, and replicates is to provide a sound scientific basis fi-om whichthe degree of certification of the resultant data can be objectively concluded. These are notmanagement decisions, but follow naturally from the results of the above QC procedures.6.1 AccuracyData accuracy is a reflection of the efficiency of the analytical procedure. It is determined byuse of spiked samples and standard reference materials or laboratory control samplesperformed at the rate of one set every 20 samples. A control chart is generated using historicallaboratory data where warning and control limits are established to assess data accuracy.The accuracy (check standards) samples will have concentration values of the mid-standard.During analysis, a minimum of 10 percent of samples must be accuracy samples. Theaccuracy samples must be staggered through the analysis, not placed one after another. After aminimum of seven accuracy samples are analyzed, the percent recovery is calculated for eachsample.OHM Project 16487SAp Sapling and Analysis plan May 1995Information herein is proprietary and confidential and to be used or released to othen only with explicit written pcmission of OHM Ram&&n&m&a Corp.6-1